A Review of Relational Machine Learning for Knowledge Graphs
By Maximilian Nickel, Kevin P. Murphy, Volker Tresp, Evgeniy Gabrilovich. 2015
1,743 citations recorded by the source corpus.
Relational machine learning studies methods for the statistical analysis of relational, or graph-structured, data. In this paper, we provide a review of how such statistical models can be “trained” on large knowledge graphs, and then used to predict new facts about the world (which is equivalent to predicting new edges in the graph). In particular, we discuss two fundamentally different kinds of statistical relational models, both of which can scale to massive data sets. The first is based on latent feature models such as tensor factorization and multiway neural networks. The second is based on mining observable patterns in the graph. We also show how to combine these latent and observable models to get improved modeling power at decreased computational cost. Finally, we discuss how such statistical models of graphs can be combined with text-based information extraction methods for automatically constructing knowledge graphs from the Web. To this end, we also discuss Google's knowledge vault project as an example of such combination.
Subfields: Artificial Intelligence, Computer Vision and Pattern Recognition, Knowledge Graphs & Representation
How researchers see this paper
The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.
In Artificial Intelligence
- Mohit Bansal
- Yoshua Bengio
- Percy Liang
- Jianfeng Gao
- Luke S. Zettlemoyer
- Sebastian Riedel
- Quoc V. Le
- Jure Leskovec
In Computer Vision and Pattern Recognition
- Luc Van Gool
- Dacheng Tao
- Trevor Darrell
- Xiaogang Wang
- Bernt Schiele
- Yi Yang
- Jitendra Malik
- Ming-Hsuan Yang
In Knowledge Graphs & Representation
- William Yang Wang
- Partha P. Talukdar
- Sebastian Riedel
- William L. Hamilton
- Avishek Joey Bose
- Seyed Mehran Kazemi
- Tim Rocktaschel
- Timothée Lacroix
Citations in the corpus
- Language Models as Knowledge Bases? (2019, methodology): “This is similar to ranking-based metrics from the knowledge base completion literature (Bordes et al., 2013; Nickel et al., 2016).”
- Neural Collaborative Filtering (2017, methodology): “Many relational machine learning methods have been devised [24].”
- Compositional Fairness Constraints for Graph Embeddings (2019, background): “, knowledge graph) setting, where embeddings are learned for both entities/nodes and relations and where the primary goal is to predict (unobserved) relations between entities [17], a general setting that encompasses both link prediction and basic recommender systems [4].”
- Representation Learning on Graphs: Methods and Applications (2017, background): “…work, which we do not review in detail here—including latent space models of social networks [32], embedding methods for statistical relational learning [42], manifold learning algorithms [37], and geometric deep learning [7]—all of which involve representation learning with graph-structured data.”
- Complex Embeddings for Simple Link Prediction (2016, methodology): “However, due to time constraints and the complexity of an efficient implementation of HolE, we record the original results for HolE as reported in Nickel et al. (2016b).”
- Representing Text for Joint Embedding of Text and Knowledge Bases (2015, methodology): “We begin by introducing notation to define the task, largely following the terminology in Nickel et al. (2015).”
- Knowledge Graph Embedding via Dynamic Mapping Matrix (2015, background): “the place of birth attribute is missing for 71% of all people included in FreeBase (Nickel et al. 2015).”
- Holographic Embeddings of Knowledge Graphs (2015, background): “(15), we are estimating a probability distribution over possible states of the knowledge graph which allows us to predict the probability of any possible triple in the graph (Nickel et al. 2015).”
- InteractE: Improving Convolution-based Knowledge Graph Embeddings by Increasing Feature Interactions (2019, background): “A survey of all variants of existing KG embedding techniques can be found in (Nickel et al. 2016; Wang et al. 2017).”
- A Survey on Knowledge Graphs: Representation, Acquisition, and Applications (2020, background): “Previous survey papers on knowledge graphs mainly focus on statistical relational learning [9], knowledge graph refinement [6], Chinese knowledge graph construction [10], KGE [8] or KRL [11].”
- Answering Visual-Relational Queries in Web-Extracted Knowledge Graphs (2017, methodology): “ImageGraph is a visual-relational KG whose relational structure is based on Freebase [Bollacker et al., 2008] and, more specifically, on FB15k, a subset of FreeBase and a popular benchmark data set [Nickel et al., 2016a].”
- Differentiable Reasoning on Large Knowledge Bases and Natural Language (2019, background): “Negative examples are obtained via a corruption process, denoted by corrupt ( · ) , by modifying the subject and object of triples in the KB (Nickel et al. 2016): NTPs can also learn interpretable rules .”
- A Survey of Fake News (2020, methodology): “To review these techniques, the following definition presents a unified standard representation of knowledge is first presented that can be automatically processed by machines and has been widely adopted in related studies [Nickel et al. 2016].”
- Learning Relational Representations with Auto-encoding Logic Programs (2019, background): “The most prominent paradigm in merging SRL and DL are (knowledge) graph embeddings [Nickel et al. , 2016; Hamilton et al. , 2017].”
- Observed versus latent features for knowledge base and text inference (2015, methodology): “One could use a margin-based loss function as used in several approaches (Nickel et al., 2015).”
- Lifted Rule Injection for Relation Embeddings (2016, methodology): “Current successful methods for automated knowledge base construction tasks heavily rely on learned distributed vector representations (Nickel et al., 2012; Riedel et al., 2013; Socher et al., 2013; Chang et al., 2014; Neelakantan et al., 2015; Toutanova et al., 2015; Nickel et al., 2015; Verga et al”